Charge pumps are well known in the art as an on-chip voltage regulator capable of providing a voltage more positive than the most positive external power supply voltage and/or negative voltage in the absence of a negative power supply voltage. The pump voltage is provided by a charge stored on a charge capacitor. The advantages of charge pumps are also well known in the art, for example, such as providing a bias voltage for the substrate of an integrated circuit or n-type and p-type wells, or for providing greater output voltage swings.
Many types of integrated circuit memories require several different power levels for operation. Some of these power levels exceed the available potential range of the external power supplies used to power the circuit. For example, access transistors connected to dynamic memory cells typically use a pumped voltage (Vccp) to drive their gates above the most positive power supply voltage. The Vccp is typically provided by a charge pump and is used to allow a complete charge to be written to the memory cell. If a lower voltage were used as the gate potential, such as the supply voltage Vcc, a threshold voltage (Vt) would be lost between the source and drain, such that a full source voltage could not reach the drain. Further, a charge pump is typically employed in non-volatile memory systems, such as flash memory systems, to provide voltages required for programming and erasing memory cells in a non-volatile memory.
Most charge pumps use an oscillator circuit, such as a ring oscillator, to provide a square wave or pulse train having voltage swings typically between ground and the most positive external supply voltage, Vcc. The pumped voltage level is partially controlled by the ring oscillator. That is, the pumped charge is generated when the ring oscillator cycles high. The ring oscillator, therefore, has an active half-cycle and an inactive half-cycle. A charge capacitor is typically pre-charged during the inactive half-cycle. The capacitor charge is then pumped to a higher level by charge sharing with another capacitor when the ring oscillator transitions to the active half-cycle. To reduce the amount of inactive time, charge pumps can have more than one phase where each phase operates on either the high or low transition of the ring oscillator.
A charge pump typically includes several pump stages which each include a pump capacitor which is charged and discharged during a clock cycle of the charge pump. A stage diode is coupled between the pump capacitor and the stage input voltage to prevent discharging of the pump capacitor prior to the pump capacitor having additional charge placed on it. Thus, a charge pump progressively stores more charge on the capacitor component of each stage, and several such stages being placed together in the charge pump produce an increasing voltage level. In integrated circuits, the diode and the capacitor are typically formed from properly configured transistors.
One or more clock signals typically trigger the charging of the pump capacitors. A typical clock signal has a clock frequency with a time period somewhat less than the discharge time of the pump capacitors. In one such design, two clock signals having opposite phase trigger the charging of alternate stages of a multi-stage charge pump. The opposite phase clock signals in this design permit increasing the amount of charge which can be placed on the pump capacitors.
A conventional pump circuit 20 is illustrated generally in schematic diagram form in FIG. 1A. Pump circuit 20 receives an input supply voltage on a line 22. A first clock signal Ph1 is received on a line 24 and a second clock signal Ph2 having the opposite phase of clock signal Ph1 is received on a line 26. FIG. 1B illustrates typical square wave clock signals Ph1 and Ph2 in timing diagram form. As illustrated in FIG. 1B, clock signals Ph1 and Ph2 have opposite phases and amplitudes corresponding to the amplitude to the input supply voltage on line 22. The input supply voltage on line 22 provides the initial supply of charge for pump circuit 20.
Pump circuit 20 includes N pump stages, as represented by a first pump stage 28 and a second pump stage 30. Clock signal Ph1 is coupled to the first pump stage 28 and clock signal Ph2 is coupled to the second pump stage 30. Similarly, clock signals Ph1 and Ph2 are alternately coupled to the remaining N-2 pump stages of pump circuit 20. Each of the N pump stages comprise a stage transistor figured to function as a diode, such as indicated at 32a for first pump stage 28 and at 32b for second pump stage 30. Each pump stage also includes a pump capacitor, such as indicated at 34a for pump stage 28 and at 34b for pump stage 30. Stage transistors 32a and 32b are typically configured to act as diodes by connecting the gate of the transistor to its drain. Pump capacitor 34a is coupled between clock signal Ph1 and the first stage output node VS1 and pump capacitor 34b is coupled between clock signal Ph2 and the second stage output node VS2.
The N pump stages are connected in a serial manner to eventually produce an Nth stage output voltage on a line 40. The Nth stage output voltage is provide to an output transistor 42. Output transistor 42 is configured to act as a diode by connecting its gate to its drain and provides a pump output voltage on a line 44.
A capacitive load 50 is coupled to the charge pump via diode connected transistor 42. The load is shown as a capacitor to emphasize the essentially capacitive nature of the load. A disadvantage of using the diode coupled transistor 42 as an isolation device is a decrease load voltage due to a voltage drop across the transistor. Under a back bias condition, a voltage drop of 2-3 volts can be experienced. As a result, an integrated circuit charge pump must be designed to produce a higher voltage on its last stage than the resultant voltage desired on the load. Because of processing technology constraints, such as junction breakdown, designing a charge pump to produce a higher output voltage is not always possible.
For the reasons stated above, and for other reasons stated below which will become apparent to those skilled in the art upon reading and understanding the present specification, there is a need in the art for a coupling circuit which allows a charge pump to be designed without the need for an elevated output voltage.